On Mon, 6 Aug 2018 at 06:09, Sankarshan Mukhopadhyay <sankarshan.mukhopadhyay@xxxxxxxxx> wrote:
On Mon, Aug 6, 2018 at 5:17 AM, Amye Scavarda <amye@xxxxxxxxxx> wrote:
>
>
> On Sun, Aug 5, 2018 at 3:24 PM Shyam Ranganathan <srangana@xxxxxxxxxx>
> wrote:
>>
>> On 07/31/2018 07:16 AM, Shyam Ranganathan wrote:
>> > On 07/30/2018 03:21 PM, Shyam Ranganathan wrote:
>> >> On 07/24/2018 03:12 PM, Shyam Ranganathan wrote:
>> >>> 1) master branch health checks (weekly, till branching)
>> >>> - Expect every Monday a status update on various tests runs
>> >> See https://build.gluster.org/job/nightly-master/ for a report on
>> >> various nightly and periodic jobs on master.
>> > Thinking aloud, we may have to stop merges to master to get these test
>> > failures addressed at the earliest and to continue maintaining them
>> > GREEN for the health of the branch.
>> >
>> > I would give the above a week, before we lockdown the branch to fix the
>> > failures.
>> >
>> > Let's try and get line-coverage and nightly regression tests addressed
>> > this week (leaving mux-regression open), and if addressed not lock the
>> > branch down.
>> >
>>
>> Health on master as of the last nightly run [4] is still the same.
>>
>> Potential patches that rectify the situation (as in [1]) are bunched in
>> a patch [2] that Atin and myself have put through several regressions
>> (mux, normal and line coverage) and these have also not passed.
>>
>> Till we rectify the situation we are locking down master branch commit
>> rights to the following people, Amar, Atin, Shyam, Vijay.
>>
>> The intention is to stabilize master and not add more patches that my
>> destabilize it.
>>
>> Test cases that are tracked as failures and need action are present here
>> [3].
>>
>> @Nigel, request you to apply the commit rights change as you see this
>> mail and let the list know regarding the same as well.
>>
>> Thanks,
>> Shyam
>>
>> [1] Patches that address regression failures:
>> https://review.gluster.org/#/q/starredby:srangana%2540redhat.com
>>
>> [2] Bunched up patch against which regressions were run:
>> https://review.gluster.org/#/c/20637
>>
>> [3] Failing tests list:
>>
>> https://docs.google.com/spreadsheets/d/1IF9GhpKah4bto19RQLr0y_Kkw26E_-crKALHSaSjZMQ/edit?usp=sharing
>>
>> [4] Nightly run dashboard: https://build.gluster.org/job/nightly-master/
>
> Locking master is fine, this seems like there's been ample notice and
> conversation.
> Do we have test criteria to indicate when we're unlocking master? X amount
> of tests passing, Y amount of bugs?
The "till we rectify" might just include 3 days of the entire set of
tests passing - thinking out loud here.
3 days = 3 nightly regressions isn’t enough as most failures are spurious in nature (IMHO). What Shyam and I are doing is retriggering various regressions on top of patch [2] . We’re looking for atleast 10 iterations to go through with out any tests retry and failures.
--
sankarshan mukhopadhyay
<https://about.me/sankarshan.mukhopadhyay>
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-devel
--Atin
_______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx https://lists.gluster.org/mailman/listinfo/gluster-devel
